Cascade: crowdsourcing taxonomy creation

By Lydia B. Chilton, Greg Little, D. Edge, Daniel S. Weld, James A. Landay. 2013

273 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Computer Science Applications, Computer Vision and Pattern Recognition, Knowledge Graphs & Representation, Human-Computer Interaction

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Computer Science Applications

In Computer Vision and Pattern Recognition

In Knowledge Graphs & Representation

In Human-Computer Interaction

Citations in the corpus

Mechanical Novel: Crowdsourcing Complex Work through Reflection and Revision (2016, methodology): “Other techniques help workers maintain global consistency: in classiﬁcation tasks, context regarding the taxonomy developed so far is provided to workers as they arrive to complete tasks in order to allow workers to consider existing categories as they classify items [9, 3].”
HumorTools : A Microtask Workflow for Writing News Satire (2015, background): “Various workflows have been proposed to cope with the variability of human output, including iterative improvement [18], findfind-verify[3], and suggestion-and-test[7].”
Communicating Context to the Crowd for Complex Writing Tasks (2017, background): “With the growth of these complex crowd-powered systems, suboptimal work is sometimes performed because individual workers lack global awareness, for instance during taxonomy creation [5].”
Bootstrapping Fine-Grained Classifiers : Active Learning with a Crowd in the Loop (2013, result): “Similar research into parallelizing crowd micro-tasks has recently been described by [13] [4] [7].”
Artificial Intelligence and Collective Intelligence (2014, methodology): “These include computing the max from a set of items (Venetis et al. 2012), multiple-choice tasks (Sun and Dance 2012), soliciting translations (Sun, Roy, and Little 2011), writing image descriptions (Little et al. 2010), and taxonomy generation (Chilton et al. 2013).”
Context Trees: Crowdsourcing Global Understanding from Local Views (2014, background): “Instead, they aggregate local contributions algorithmically so that no crowd member ever needs a higherlevel view, as for example with clustering (Chilton et al. 2013; Bragg, Mausam, and Weld 2013; Tamuz et al. 2011; Gomes et al. 2011).”
Crowdsourced Data Management: Industry and Academic Perspectives (2015, unknown): “[55], and Bragg et al.”
No Workflow Can Ever Be Enough (2017, methodology): “Challenges include poor early results that derail later stages [44], uncoordinated contributions that make inconsistent changes [9, 33], and workers who do not have sufficient global context to make effective decisions [16, 30, 36].”
Examining Crowd Work and Gig Work Through the Historical Lens of Piecework (2017, background): “Third, data analysis tasks such as clustering [34], categorization [10], and outlining [99] are possible, but there is no general solution for sense–making.”
Revolt: Collaborative Crowdsourcing for Labeling Machine Learning Datasets (2017, methodology): “For example, categorizing [12, 4] or create tax-onomies [15] for a set of documents.”
Human Uncertainty Makes Classification More Robust (2019, background): “The label hierarchies used to aid recognition can be manually constructed [6, 3], derived from linguistic knowledge bases [10, 9], or learned automatically [14, 19].”
Flock: Hybrid Crowd-Machine Learning Classifiers (2015, background): “[7]), we used six different problem domains.”
Productivity Decomposed: Getting Big Things Done with Little Microtasks (2016, background): “On crowd platforms, microtasks are increasingly being composed to accomplish complex tasks that are not obviously achievable via standalone microtasks, such as taxonomy creation [7], itinerary planning [15], and writing [3].”